dali-mcp 0.1.0__tar.gz

dali_mcp-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,40 @@
+name: Publish to PyPI
+
+on:
+  release:
+    types: [published]
+  workflow_dispatch: {}
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.x"
+      - name: Build package
+        run: |
+          pip install build
+          python -m build
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+  publish:
+    needs: build
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      id-token: write
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

dali_mcp-0.1.0/.gitignore

@@ -0,0 +1,10 @@
+__pycache__/
+*.py[cod]
+*.egg-info/
+dist/
+build/
+.venv/
+venv/
+.env
+*.env
+.DS_Store

dali_mcp-0.1.0/Dockerfile

@@ -0,0 +1,15 @@
+FROM python:3.12-slim
+
+WORKDIR /app
+
+COPY pyproject.toml .
+RUN pip install --no-cache-dir "fastmcp>=0.1.0" "httpx>=0.27.0" hatchling
+
+COPY dali/ ./dali/
+
+# DALI_TOKEN required — get yours at dali.getlulu.dev (free GitHub OAuth login)
+ENV DALI_TRANSPORT=http
+
+EXPOSE 8080
+
+CMD ["python", "-m", "dali.server"]

dali_mcp-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Lulu
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

dali_mcp-0.1.0/PKG-INFO

@@ -0,0 +1,225 @@
+Metadata-Version: 2.4
+Name: dali-mcp
+Version: 0.1.0
+Summary: Dali by Lulu — creative intelligence MCP. Score your prompt before you spend the credit.
+Project-URL: Homepage, https://dali.getlulu.dev
+Project-URL: Repository, https://github.com/Lulu-The-Narwhal/dali-mcp
+License: MIT
+License-File: LICENSE
+Keywords: ai,creative,flux,higgsfield,image,lulu,mcp,midjourney,prompts,veo,video
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Requires-Python: >=3.10
+Requires-Dist: fastmcp>=0.1.0
+Requires-Dist: httpx>=0.27.0
+Description-Content-Type: text/markdown
+
+<!-- mcp-name: io.github.Lulu-The-Narwhal/dali -->
+# Dali by Lulu
+
+<p align="center">
+  <img src="assets/dali-lulu-scorer.png" alt="Dali by Lulu — creative intelligence MCP" width="100%">
+</p>
+
+<p align="center">
+  <a href="https://dali.getlulu.dev"><strong>dali.getlulu.dev</strong></a> &nbsp;·&nbsp;
+  <a href="https://dali.getlulu.dev/#install">Install</a> &nbsp;·&nbsp;
+  <a href="https://dali.getlulu.dev/dashboard">Live stats</a> &nbsp;·&nbsp;
+  <a href="https://getlulu.dev">Lulu</a>
+</p>
+
+<p align="center">
+  <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="MIT License"></a>
+  <a href="https://www.python.org/downloads/"><img src="https://img.shields.io/badge/python-3.10%2B-blue.svg" alt="Python 3.10+"></a>
+  <a href="https://modelcontextprotocol.io"><img src="https://img.shields.io/badge/MCP-server-6b5bff.svg" alt="MCP Server"></a>
+  <a href="https://dali.getlulu.dev/dashboard"><img src="https://img.shields.io/badge/status-live-brightgreen.svg" alt="Live"></a>
+</p>
+
+---
+
+**The prediction MCP that helps you avoid the AI generation tax.**
+
+Most AI generation failures are prompt failures. You can't tell the difference until after you've burned the token. Dali scores your prompt *before* you generate — so you never waste a credit on a bad prompt again. Every wasted generation has a real cost (a Seedance retry is ~$6) — [the live dashboard](https://dali.getlulu.dev/dashboard) tracks what the community has saved by catching bad prompts before they burned a credit.
+
+```
+You: "make a video ad for our glass serum bottle"
+
+dali::score_prompt(prompt, "veo3")
+→ 8/100  Grade: F
+→ no camera move · no motion · no lighting · 8 words
+→ Verdict: Generic stock footage guaranteed. Enhance first.
+
+dali::enhance_prompt(prompt, "veo3")
+→ Returns a rewrite brief — YOUR LLM writes the enhanced prompt:
+
+  ① lead with camera — Veo 3's #1 lever: "Slow dolly", "Orbital push"
+  ② describe physics: "a drop falls", "liquid ripples", "glass refracts"
+  ③ lighting type + quality: "warm backlight", "rim-lit edges"
+  ↳ [Camera]. [Subject + motion]. [Lighting]. [Mood]. [No text.]
+
+✦ Claude rewrites using the brief:
+
+  "Slow orbital push around a glass serum bottle on white marble. A single
+   amber drop falls in extreme slow motion, catching warm backlight. Macro:
+   liquid gold ripples outward from impact. Rim-lit edges, soft studio
+   diffusion. Premium, clinical. No text."
+
+dali::score_prompt(enhanced, "veo3")
+→ 91/100  Grade: A  ✓ Safe to generate.
+```
+
+---
+
+## Contents
+
+- [Install](#install)
+- [Tools](#tools)
+- [Supported models](#supported-models)
+- [Platform supersets](#platform-supersets)
+- [Why model-specific?](#why-model-specific)
+- [MCP resources](#mcp-resources)
+- [Contributing](#contributing)
+
+---
+
+## Install
+
+**Hosted MCP — connect once, scores every prompt:**
+
+```bash
+# Claude Code
+claude mcp add --transport http dali https://dali.getlulu.dev/mcp
+```
+
+```json
+// Cursor / Windsurf — .cursor/mcp.json or windsurf settings
+{
+  "mcpServers": {
+    "dali": { "url": "https://dali.getlulu.dev/mcp" }
+  }
+}
+```
+
+→ **[Full install guide with all clients](https://dali.getlulu.dev/#install)**
+
+**Self-hosted — local, no auth required:**
+
+```bash
+# PyPI package is not published yet — install directly from source for now
+pip install git+https://github.com/Lulu-The-Narwhal/dali-mcp
+claude mcp add dali -- python -m dali.server
+```
+
+---
+
+## Tools
+
+| Tool | What it does |
+|------|-------------|
+| `score_prompt(prompt, model)` | Grade 0–100, letter grade, per-dimension breakdown, what's missing, verdict |
+| `enhance_prompt(prompt, model)` | Returns a structured rewrite brief — YOUR LLM writes the enhanced prompt using it |
+| `analyze_intent(prompt)` | Parse dimensions: camera, motion, lighting, style, mood, gaps |
+| `creative_patterns(model)` | Community top patterns for this model from the graph brain |
+| `community_benchmark(prompt, model)` | Compare your prompt against community top scorers |
+| `my_story()` | Your scoring history, model stats, grade distribution |
+| `list_models()` | All supported models with medium and core strength |
+
+---
+
+## Supported models
+
+### Video
+
+| Model | Platforms | Best for | Prompt style |
+|-------|-----------|----------|--------------|
+| `veo3` | Higgsfield, Google AI Studio (`veo-3.1-generate-preview`), Runway | Cinematic brand films, narrative ads, photorealistic motion | Camera move → Subject → Action → Location → Lighting → Mood |
+| `seedance` | Higgsfield, fal.ai (`bytedance/seedance-2.0`) | UGC, social-native content, TikTok/Reels performance ads | Natural language, motion-first, authentic feel |
+| `kling` | Higgsfield (`kling3`), Kling.ai (`kling-v3-text-to-video`) | Character animation, product showcases, facial performance | Scene → Characters → Action → Camera → Style; multi-shot labels |
+| `runway` | Runway (`gen4_turbo`) | VFX, character performance, cinematic motion | Motion-first — describe what moves, not what exists |
+| `wan` | fal.ai (`fal-ai/wan/v2.7/text-to-video`) | 4K, 20-second clips, native audio, open-source workflows | Scene → Motion → Sound → Duration → Mood |
+| `minimax` | fal.ai (`fal-ai/minimax/hailuo-02/pro/text-to-video`) | Cinematic storytelling, character animation | Natural language + `[camera movement]` bracket syntax |
+| `higgsfield` | Higgsfield (native model) | Physics-driven motion — cloth, hair, fluid, particles | Describe materials in motion, not motion abstractly |
+
+> **Sora 2** (OpenAI): API shutdown September 24, 2026. Do not build new dependencies on it — use Runway or Kling instead.
+
+### Image
+
+| Model | Platforms | Best for | Prompt style |
+|-------|-----------|----------|--------------|
+| `flux` | BFL API (`flux-pro-v1.1`), fal.ai, Replicate | Photorealism, technical photography, product shots | 30–80 words; camera body + lens specs; front-load subject |
+| `midjourney` | Midjourney (v8.1) | Artistic depth, editorial, stylized illustration | Prose + params appended: `--ar 16:9 --s 300 --v 8.1 --style raw` |
+| `ideogram` | Ideogram API (`V_4`), fal.ai | Typography, logos, text-in-image, graphic design | Describe text exactly in quotes inside the prompt |
+| `firefly` | Adobe Firefly 5 (enterprise) | IP-indemnified commercial assets, 4MP brand content | Natural language + `contentClass` and `style.presets` API params |
+
+> **Imagen 4** (Google): deprecated — use `gemini-3.5-flash` with image output. Dali still scores legacy Imagen prompts via the `imagen` model key but don't build new things on it.
+
+---
+
+## Platform supersets
+
+**Higgsfield** and **Runway** are aggregator platforms — they proxy multiple underlying models under one API. The model you pick matters more than the platform name:
+
+| Platform | Model selector | Underlying model |
+|----------|---------------|-----------------|
+| Higgsfield | `veo3` | Google Veo 3.1 |
+| Higgsfield | `seedance` | ByteDance Seedance 2.0 |
+| Higgsfield | `kling3` | Kling 3 |
+| Higgsfield | `wan2-7` | Wan 2.7 |
+| Higgsfield | `image2video` | Higgsfield native |
+| Runway | `veo3` | Google Veo 3.1 |
+| Runway | `gen4_turbo` | Runway Gen 4.5 |
+| Runway | `seedance` | ByteDance Seedance 2.0 |
+
+Dali scores for the **underlying model's native prompt language**, not the platform wrapper. Pass the model name (`veo3`, `kling`, `seedance`…), not the platform name.
+
+---
+
+## Why model-specific?
+
+Generic prompt optimizers don't know that:
+- **Veo 3.1** needs camera movement specified above everything else
+- **Kling 3** supports multi-shot scene labels natively in the prompt
+- **Flux** responds to camera body and lens names like a photographer (`"Sony A7 IV, 85mm f/1.4"`)
+- **Midjourney V8.1** reads prose + parameters, not keyword lists
+- **Higgsfield** simulates physics — you describe materials in motion, not motion abstractly
+- **Minimax** uses `[Pan left]` bracket syntax for camera moves — plain text camera commands are ignored
+- **Ideogram V4** needs text quoted exactly in the prompt for typography accuracy
+- **Wan 2.7** generates native audio — include sound descriptions alongside visuals
+
+Dali has a separate scoring rubric and rewrite brief for each model. Your LLM does the creative rewriting — Dali provides the intelligence.
+
+---
+
+## MCP resources
+
+```
+creative://guide/veo3       → Veo 3.1 camera language guide
+creative://guide/seedance   → Seedance UGC motion guide
+creative://guide/kling      → Kling multi-shot + expression guide
+creative://guide/runway     → Runway motion-first guide
+creative://guide/wan        → Wan 2.7 audio + motion guide
+creative://guide/minimax    → Minimax bracket camera guide
+creative://guide/higgsfield → Higgsfield physics-motion guide
+creative://guide/sora       → Sora 2 guide (API shutdown Sep 24, 2026)
+creative://guide/flux       → Flux photography brief guide
+creative://guide/midjourney → Midjourney V8.1 + parameters guide
+creative://guide/ideogram   → Ideogram V4 typography guide
+creative://guide/firefly    → Firefly 5 commercial content guide
+creative://guide/imagen     → Imagen 4 guide (deprecated Aug 17, 2026)
+creative://models           → All models overview
+```
+
+---
+
+## Contributing
+
+Model guides live in `dali/data/guides/{model}.json` on the hosted server. Found practitioner patterns that consistently produce high-grade results? Open an issue with the model, the pattern, and a sample prompt + result. The best contributions come from Reddit, Discord, and YouTube — real practitioners, not official docs.
+
+→ **[Prompt best practices by model](docs/best-practices.md)** — cheat sheets, do/don't tables, top patterns per model
+→ **[Dali creative flow skill](skills/dali-creative-flow.md)** — install this skill so your LLM follows the score → enhance → generate workflow automatically
+
+---
+
+[MIT License](LICENSE) · Built by [Lulu](https://getlulu.dev) · [dali.getlulu.dev](https://dali.getlulu.dev)

dali_mcp-0.1.0/README.md

@@ -0,0 +1,207 @@
+<!-- mcp-name: io.github.Lulu-The-Narwhal/dali -->
+# Dali by Lulu
+
+<p align="center">
+  <img src="assets/dali-lulu-scorer.png" alt="Dali by Lulu — creative intelligence MCP" width="100%">
+</p>
+
+<p align="center">
+  <a href="https://dali.getlulu.dev"><strong>dali.getlulu.dev</strong></a> &nbsp;·&nbsp;
+  <a href="https://dali.getlulu.dev/#install">Install</a> &nbsp;·&nbsp;
+  <a href="https://dali.getlulu.dev/dashboard">Live stats</a> &nbsp;·&nbsp;
+  <a href="https://getlulu.dev">Lulu</a>
+</p>
+
+<p align="center">
+  <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="MIT License"></a>
+  <a href="https://www.python.org/downloads/"><img src="https://img.shields.io/badge/python-3.10%2B-blue.svg" alt="Python 3.10+"></a>
+  <a href="https://modelcontextprotocol.io"><img src="https://img.shields.io/badge/MCP-server-6b5bff.svg" alt="MCP Server"></a>
+  <a href="https://dali.getlulu.dev/dashboard"><img src="https://img.shields.io/badge/status-live-brightgreen.svg" alt="Live"></a>
+</p>
+
+---
+
+**The prediction MCP that helps you avoid the AI generation tax.**
+
+Most AI generation failures are prompt failures. You can't tell the difference until after you've burned the token. Dali scores your prompt *before* you generate — so you never waste a credit on a bad prompt again. Every wasted generation has a real cost (a Seedance retry is ~$6) — [the live dashboard](https://dali.getlulu.dev/dashboard) tracks what the community has saved by catching bad prompts before they burned a credit.
+
+```
+You: "make a video ad for our glass serum bottle"
+
+dali::score_prompt(prompt, "veo3")
+→ 8/100  Grade: F
+→ no camera move · no motion · no lighting · 8 words
+→ Verdict: Generic stock footage guaranteed. Enhance first.
+
+dali::enhance_prompt(prompt, "veo3")
+→ Returns a rewrite brief — YOUR LLM writes the enhanced prompt:
+
+  ① lead with camera — Veo 3's #1 lever: "Slow dolly", "Orbital push"
+  ② describe physics: "a drop falls", "liquid ripples", "glass refracts"
+  ③ lighting type + quality: "warm backlight", "rim-lit edges"
+  ↳ [Camera]. [Subject + motion]. [Lighting]. [Mood]. [No text.]
+
+✦ Claude rewrites using the brief:
+
+  "Slow orbital push around a glass serum bottle on white marble. A single
+   amber drop falls in extreme slow motion, catching warm backlight. Macro:
+   liquid gold ripples outward from impact. Rim-lit edges, soft studio
+   diffusion. Premium, clinical. No text."
+
+dali::score_prompt(enhanced, "veo3")
+→ 91/100  Grade: A  ✓ Safe to generate.
+```
+
+---
+
+## Contents
+
+- [Install](#install)
+- [Tools](#tools)
+- [Supported models](#supported-models)
+- [Platform supersets](#platform-supersets)
+- [Why model-specific?](#why-model-specific)
+- [MCP resources](#mcp-resources)
+- [Contributing](#contributing)
+
+---
+
+## Install
+
+**Hosted MCP — connect once, scores every prompt:**
+
+```bash
+# Claude Code
+claude mcp add --transport http dali https://dali.getlulu.dev/mcp
+```
+
+```json
+// Cursor / Windsurf — .cursor/mcp.json or windsurf settings
+{
+  "mcpServers": {
+    "dali": { "url": "https://dali.getlulu.dev/mcp" }
+  }
+}
+```
+
+→ **[Full install guide with all clients](https://dali.getlulu.dev/#install)**
+
+**Self-hosted — local, no auth required:**
+
+```bash
+# PyPI package is not published yet — install directly from source for now
+pip install git+https://github.com/Lulu-The-Narwhal/dali-mcp
+claude mcp add dali -- python -m dali.server
+```
+
+---
+
+## Tools
+
+| Tool | What it does |
+|------|-------------|
+| `score_prompt(prompt, model)` | Grade 0–100, letter grade, per-dimension breakdown, what's missing, verdict |
+| `enhance_prompt(prompt, model)` | Returns a structured rewrite brief — YOUR LLM writes the enhanced prompt using it |
+| `analyze_intent(prompt)` | Parse dimensions: camera, motion, lighting, style, mood, gaps |
+| `creative_patterns(model)` | Community top patterns for this model from the graph brain |
+| `community_benchmark(prompt, model)` | Compare your prompt against community top scorers |
+| `my_story()` | Your scoring history, model stats, grade distribution |
+| `list_models()` | All supported models with medium and core strength |
+
+---
+
+## Supported models
+
+### Video
+
+| Model | Platforms | Best for | Prompt style |
+|-------|-----------|----------|--------------|
+| `veo3` | Higgsfield, Google AI Studio (`veo-3.1-generate-preview`), Runway | Cinematic brand films, narrative ads, photorealistic motion | Camera move → Subject → Action → Location → Lighting → Mood |
+| `seedance` | Higgsfield, fal.ai (`bytedance/seedance-2.0`) | UGC, social-native content, TikTok/Reels performance ads | Natural language, motion-first, authentic feel |
+| `kling` | Higgsfield (`kling3`), Kling.ai (`kling-v3-text-to-video`) | Character animation, product showcases, facial performance | Scene → Characters → Action → Camera → Style; multi-shot labels |
+| `runway` | Runway (`gen4_turbo`) | VFX, character performance, cinematic motion | Motion-first — describe what moves, not what exists |
+| `wan` | fal.ai (`fal-ai/wan/v2.7/text-to-video`) | 4K, 20-second clips, native audio, open-source workflows | Scene → Motion → Sound → Duration → Mood |
+| `minimax` | fal.ai (`fal-ai/minimax/hailuo-02/pro/text-to-video`) | Cinematic storytelling, character animation | Natural language + `[camera movement]` bracket syntax |
+| `higgsfield` | Higgsfield (native model) | Physics-driven motion — cloth, hair, fluid, particles | Describe materials in motion, not motion abstractly |
+
+> **Sora 2** (OpenAI): API shutdown September 24, 2026. Do not build new dependencies on it — use Runway or Kling instead.
+
+### Image
+
+| Model | Platforms | Best for | Prompt style |
+|-------|-----------|----------|--------------|
+| `flux` | BFL API (`flux-pro-v1.1`), fal.ai, Replicate | Photorealism, technical photography, product shots | 30–80 words; camera body + lens specs; front-load subject |
+| `midjourney` | Midjourney (v8.1) | Artistic depth, editorial, stylized illustration | Prose + params appended: `--ar 16:9 --s 300 --v 8.1 --style raw` |
+| `ideogram` | Ideogram API (`V_4`), fal.ai | Typography, logos, text-in-image, graphic design | Describe text exactly in quotes inside the prompt |
+| `firefly` | Adobe Firefly 5 (enterprise) | IP-indemnified commercial assets, 4MP brand content | Natural language + `contentClass` and `style.presets` API params |
+
+> **Imagen 4** (Google): deprecated — use `gemini-3.5-flash` with image output. Dali still scores legacy Imagen prompts via the `imagen` model key but don't build new things on it.
+
+---
+
+## Platform supersets
+
+**Higgsfield** and **Runway** are aggregator platforms — they proxy multiple underlying models under one API. The model you pick matters more than the platform name:
+
+| Platform | Model selector | Underlying model |
+|----------|---------------|-----------------|
+| Higgsfield | `veo3` | Google Veo 3.1 |
+| Higgsfield | `seedance` | ByteDance Seedance 2.0 |
+| Higgsfield | `kling3` | Kling 3 |
+| Higgsfield | `wan2-7` | Wan 2.7 |
+| Higgsfield | `image2video` | Higgsfield native |
+| Runway | `veo3` | Google Veo 3.1 |
+| Runway | `gen4_turbo` | Runway Gen 4.5 |
+| Runway | `seedance` | ByteDance Seedance 2.0 |
+
+Dali scores for the **underlying model's native prompt language**, not the platform wrapper. Pass the model name (`veo3`, `kling`, `seedance`…), not the platform name.
+
+---
+
+## Why model-specific?
+
+Generic prompt optimizers don't know that:
+- **Veo 3.1** needs camera movement specified above everything else
+- **Kling 3** supports multi-shot scene labels natively in the prompt
+- **Flux** responds to camera body and lens names like a photographer (`"Sony A7 IV, 85mm f/1.4"`)
+- **Midjourney V8.1** reads prose + parameters, not keyword lists
+- **Higgsfield** simulates physics — you describe materials in motion, not motion abstractly
+- **Minimax** uses `[Pan left]` bracket syntax for camera moves — plain text camera commands are ignored
+- **Ideogram V4** needs text quoted exactly in the prompt for typography accuracy
+- **Wan 2.7** generates native audio — include sound descriptions alongside visuals
+
+Dali has a separate scoring rubric and rewrite brief for each model. Your LLM does the creative rewriting — Dali provides the intelligence.
+
+---
+
+## MCP resources
+
+```
+creative://guide/veo3       → Veo 3.1 camera language guide
+creative://guide/seedance   → Seedance UGC motion guide
+creative://guide/kling      → Kling multi-shot + expression guide
+creative://guide/runway     → Runway motion-first guide
+creative://guide/wan        → Wan 2.7 audio + motion guide
+creative://guide/minimax    → Minimax bracket camera guide
+creative://guide/higgsfield → Higgsfield physics-motion guide
+creative://guide/sora       → Sora 2 guide (API shutdown Sep 24, 2026)
+creative://guide/flux       → Flux photography brief guide
+creative://guide/midjourney → Midjourney V8.1 + parameters guide
+creative://guide/ideogram   → Ideogram V4 typography guide
+creative://guide/firefly    → Firefly 5 commercial content guide
+creative://guide/imagen     → Imagen 4 guide (deprecated Aug 17, 2026)
+creative://models           → All models overview
+```
+
+---
+
+## Contributing
+
+Model guides live in `dali/data/guides/{model}.json` on the hosted server. Found practitioner patterns that consistently produce high-grade results? Open an issue with the model, the pattern, and a sample prompt + result. The best contributions come from Reddit, Discord, and YouTube — real practitioners, not official docs.
+
+→ **[Prompt best practices by model](docs/best-practices.md)** — cheat sheets, do/don't tables, top patterns per model
+→ **[Dali creative flow skill](skills/dali-creative-flow.md)** — install this skill so your LLM follows the score → enhance → generate workflow automatically
+
+---
+
+[MIT License](LICENSE) · Built by [Lulu](https://getlulu.dev) · [dali.getlulu.dev](https://dali.getlulu.dev)

dali_mcp-0.1.0/cloudbuild.yaml

@@ -0,0 +1,46 @@
+steps:
+  # Build image
+  - name: 'gcr.io/cloud-builders/docker'
+    args:
+      - build
+      - -t
+      - us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:$COMMIT_SHA
+      - -t
+      - us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:latest
+      - .
+
+  # Push SHA tag
+  - name: 'gcr.io/cloud-builders/docker'
+    args: ['push', 'us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:$COMMIT_SHA']
+
+  # Push latest tag
+  - name: 'gcr.io/cloud-builders/docker'
+    args: ['push', 'us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:latest']
+
+  # Get GKE credentials
+  - name: 'gcr.io/google.com/cloudsdktool/cloud-sdk'
+    entrypoint: gcloud
+    args:
+      - container
+      - clusters
+      - get-credentials
+      - lulu-cluster
+      - --region=us-east1
+      - --project=$PROJECT_ID
+
+  # Roll dali-mcp deployment to the new image
+  - name: 'gcr.io/cloud-builders/kubectl'
+    args:
+      - set
+      - image
+      - deployment/dali-mcp
+      - dali-mcp=us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:$COMMIT_SHA
+      - -n
+      - lulu
+    env:
+      - 'CLOUDSDK_COMPUTE_REGION=us-east1'
+      - 'CLOUDSDK_CONTAINER_CLUSTER=lulu-cluster'
+
+images:
+  - us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:$COMMIT_SHA
+  - us-central1-docker.pkg.dev/$PROJECT_ID/dali-mcp/dali-mcp:latest

dali_mcp-0.1.0/dali/__init__.py

@@ -0,0 +1,3 @@
+"""Dali by Lulu — creative intelligence MCP for AI generation prompts."""
+
+__version__ = "0.1.0"

dali_mcp-0.1.0/dali/client.py

@@ -0,0 +1,50 @@
+"""
+HTTP client for the Dali hosted API at dali.getlulu.dev.
+
+All intelligence (scoring, enhancement, Memgraph graph) runs server-side.
+This module handles auth token injection and request formatting.
+
+Tokens: get yours at dali.getlulu.dev (GitHub OAuth login, free tier).
+"""
+
+from __future__ import annotations
+import os
+import httpx
+from typing import Optional
+
+DALI_API_BASE = os.environ.get("DALI_API_URL", "https://dali.getlulu.dev")
+_DEFAULT_TIMEOUT = 30.0
+
+
+def _headers(token: Optional[str] = None) -> dict:
+    h = {"Content-Type": "application/json", "User-Agent": "dali-mcp/0.3.0"}
+    tok = token or os.environ.get("DALI_TOKEN")
+    if tok:
+        h["Authorization"] = f"Bearer {tok}"
+    return h
+
+
+def call(endpoint: str, payload: dict, token: Optional[str] = None) -> dict:
+    """POST to the Dali API. Returns the parsed JSON response."""
+    url = f"{DALI_API_BASE.rstrip('/')}/{endpoint.lstrip('/')}"
+    try:
+        with httpx.Client(timeout=_DEFAULT_TIMEOUT) as http:
+            resp = http.post(url, json=payload, headers=_headers(token))
+            resp.raise_for_status()
+            return resp.json()
+    except httpx.TimeoutException:
+        return {"error": "Request timed out. Dali API may be under load — retry in a moment."}
+    except httpx.HTTPStatusError as e:
+        if e.response.status_code == 401:
+            return {
+                "error": "Unauthorized. Set DALI_TOKEN or run: dali auth login",
+                "auth_url": f"{DALI_API_BASE}/auth/github",
+            }
+        if e.response.status_code == 429:
+            return {
+                "error": "Rate limit reached. Free tier: 100 calls/month.",
+                "upgrade_url": f"{DALI_API_BASE}/pricing",
+            }
+        return {"error": f"API error {e.response.status_code}: {e.response.text[:200]}"}
+    except Exception as e:
+        return {"error": f"Could not reach Dali API: {e}"}

dali_mcp-0.1.0/dali/server.py

@@ -0,0 +1,266 @@
+"""
+Dali by Lulu — creative intelligence MCP for AI generation agents.
+
+Score your prompt before you spend the credit.
+
+This is the open-source MCP client. Tool definitions live here.
+Intelligence (scoring engine, Gemini enhancement, V3 graph brain) runs
+on the hosted server at dali.getlulu.dev.
+
+Connect (recommended):
+  claude mcp add dali --url https://dali.getlulu.dev/mcp
+
+Self-hosted (calls Dali API, requires a free token):
+  pip install dali-mcp
+  export DALI_TOKEN=<your token from dali.getlulu.dev>
+  claude mcp add dali -- python -m dali.server
+
+Tools:
+  analyze_intent       — parse raw text into structured creative intent
+  score_prompt         — score a prompt 0–100 for a specific model
+  enhance_prompt       — Gemini-powered rewrite in the model's native language
+  creative_patterns    — community graph: A-grade patterns for this model (V3)
+  community_benchmark  — how your prompt compares to community top scorers (V3)
+  my_story             — your scoring history, model stats, creative DNA
+  list_models          — all supported generation models
+
+Resources:
+  creative://guide/{model}  — full native-language guide for each model
+  creative://models         — overview of all models
+"""
+
+from __future__ import annotations
+import os
+
+try:
+    from fastmcp import FastMCP
+except ImportError:
+    raise ImportError(
+        "fastmcp not installed. Run: pip install dali-mcp\n"
+        "Or use the hosted server: https://dali.getlulu.dev/mcp"
+    )
+
+from .client import call
+
+mcp = FastMCP(
+    name="dali",
+    instructions=(
+        "Dali is creative intelligence for AI generation. Use it BEFORE generating any image or video.\n\n"
+        "PROACTIVE WORKFLOW (always follow this):\n"
+        "1. User expresses any image/video generation intent → IMMEDIATELY call score_prompt(prompt, model)\n"
+        "2. Grade A or B (score ≥ 70) → proceed with the prompt\n"
+        "3. Grade C, D, or F (score < 70) → call enhance_prompt(prompt, model) to get a rewrite brief, "
+        "then YOU write the enhanced prompt using the brief's native_language_rules + structure_template. "
+        "Then call score_prompt on your version to verify.\n"
+        "4. NEVER generate without scoring first\n\n"
+        "TOOLS:\n"
+        "- analyze_intent(prompt) — parse intent signals before scoring\n"
+        "- score_prompt(prompt, model) — grade 0–100 with what's missing\n"
+        "- enhance_prompt(prompt, model) — AI rewrite in the model's native language\n"
+        "- creative_patterns(model) — community A-grade patterns from the graph brain\n"
+        "- community_benchmark(prompt, model) — compare vs top scorers\n"
+        "- my_story() — personal scoring history and creative DNA\n"
+        "- list_models() — all supported models\n\n"
+        "SUPPORTED MODELS: veo3, seedance, kling, runway, wan, minimax, higgsfield, sora, flux, midjourney, ideogram, firefly, imagen\n"
+        "ALIASES: veo→veo3, mj→midjourney, sd→flux, gen4→runway, hailuo→minimax, wan27→wan\n\n"
+        "RESOURCES: creative://guide/{model} — full native-language guide\n\n"
+        "INSTALL GUIDE: dali.getlulu.dev/install"
+    ),
+)
+
+
+# ── Tools ──────────────────────────────────────────────────────────────────
+
+@mcp.tool()
+def analyze_intent(
+    prompt: str,
+    medium: str = "auto",
+) -> dict:
+    """
+    Parse a creative prompt into structured intent dimensions.
+
+    Returns: detected camera language, motion, lighting, style, mood signals,
+    identified gaps, and which models suit the current signals best.
+
+    Args:
+        prompt: Raw creative text (rough idea or full prompt — both work)
+        medium: "image", "video", or "auto" (default, auto-detected)
+    """
+    return call("/api/intent", {"prompt": prompt, "medium": medium})
+
+
+@mcp.tool()
+def score_prompt(
+    prompt: str,
+    model: str,
+) -> dict:
+    """
+    Score a prompt for a specific generation model (0–100).
+
+    Returns a ScoreCard: overall score, letter grade (A–F), per-dimension
+    breakdown, what's missing, detected anti-patterns, and a one-line verdict
+    on whether it's safe to generate.
+
+    Supported models: veo3, seedance, kling, runway, wan, minimax, higgsfield, sora, flux, midjourney, ideogram, firefly, imagen
+    Aliases: "veo" → veo3, "mj" → midjourney, "sd" → flux, "gen4" → runway, "hailuo" → minimax
+
+    Args:
+        prompt: The prompt to score
+        model:  Target generation model
+    """
+    return call("/api/score", {"prompt": prompt, "model": model})
+
+
+@mcp.tool()
+def enhance_prompt(
+    prompt: str,
+    model: str,
+) -> dict:
+    """
+    Rewrite a prompt using AI to score higher on the target model.
+
+    Returns a rewrite brief — YOU (the LLM) write the enhanced prompt from it.
+
+    Dali provides creative intelligence: what's missing, the model's native language rules,
+    structure template, priority fixes, and length target.
+    You provide creative execution: actually writing the better prompt.
+
+    Returns:
+      - score_before:    full ScoreCard showing current gaps
+      - rewrite_brief:   native_language_rules, structure_template, priority_fixes, length_target
+      - llm_instructions: step-by-step instructions for writing the enhanced prompt
+
+    Args:
+        prompt: The prompt to enhance
+        model:  Target generation model (veo3, seedance, kling, runway, wan, minimax, higgsfield, sora, flux, midjourney, ideogram, firefly, imagen)
+    """
+    return call("/api/enhance", {"prompt": prompt, "model": model})
+
+
+@mcp.tool()
+def creative_patterns(
+    model: str,
+    grade: str = "A",
+) -> dict:
+    """
+    Community graph intelligence: which patterns consistently produce high-grade prompts
+    for this model?
+
+    Powered by the Dali V3 graph brain — every prompt scored by every
+    Dali user contributes to this. The more community usage, the richer the signal.
+
+    Also returns: enhancement unlocks (which patterns added during enhance_prompt
+    have produced the highest score gains for this model).
+
+    Args:
+        model: Target generation model (veo3, seedance, kling, runway, wan, minimax, higgsfield, flux, midjourney, ideogram, firefly)
+        grade: Minimum grade filter — "A" (only A-grade), "B" (A+B), "C" (A+B+C)
+    """
+    return call("/api/patterns", {"model": model, "grade": grade})
+
+
+@mcp.tool()
+def community_benchmark(
+    prompt: str,
+    model: str,
+) -> dict:
+    """
+    Compare your prompt against community top scorers for this model.
+
+    Scores your prompt, then queries the Dali graph to find:
+    - Which A-grade community patterns are absent from your prompt
+    - Which enhancement patterns would give you the biggest score gain
+    - Where you sit vs the community average for this model
+
+    Args:
+        prompt: Your prompt to benchmark
+        model:  Target generation model
+    """
+    return call("/api/benchmark", {"prompt": prompt, "model": model})
+
+
+@mcp.tool()
+def my_story() -> dict:
+    """
+    Your Dali creative intelligence report.
+
+    Shows your prompt scoring history across all models:
+    - Total prompts scored + this month's count
+    - Average score and model breakdown
+    - Grade distribution (how many A, B, C, D, F scores)
+    - Your creative DNA: patterns that define your highest-scoring work (V3 graph)
+    - A personal insight on where you'd improve fastest
+
+    Requires login at dali.getlulu.dev. History is tied to your GitHub account.
+    """
+    return call("/api/story", {})
+
+
+@mcp.tool()
+def list_models() -> dict:
+    """List all supported generation models with medium, creator, and core strength."""
+    return call("/api/models", {})
+
+
+# ── Resources ──────────────────────────────────────────────────────────────
+
+@mcp.resource("creative://guide/{model}")
+def model_guide(model: str) -> str:
+    """Full native-language prompt guide for a specific generation model."""
+    result = call("/api/guide", {"model": model})
+    if "error" in result:
+        return f"Could not load guide: {result['error']}"
+    return result.get("text", f"No guide found for {model}")
+
+
+@mcp.resource("creative://models")
+def models_overview() -> str:
+    """Overview of all supported models and their strengths."""
+    result = call("/api/models", {})
+    if "error" in result:
+        return "Could not load models overview. Is DALI_TOKEN set?"
+    lines = [
+        "# Dali by Lulu — Supported Generation Models",
+        "dali.getlulu.dev | github.com/Lulu-The-Narwhal/dali-mcp",
+        "",
+    ]
+    details = result.get("details", {})
+    video = [m for m, d in details.items() if d.get("medium") == "video"]
+    image = [m for m, d in details.items() if d.get("medium") == "image"]
+    lines.append("## Video Models")
+    for m in video:
+        d = details[m]
+        lines.append(f"- **{m}** ({d.get('by', '')}): {d.get('strength', '')}")
+    lines.append("\n## Image Models")
+    for m in image:
+        d = details[m]
+        lines.append(f"- **{m}** ({d.get('by', '')}): {d.get('strength', '')}")
+    lines += [
+        "\n## Tools",
+        "- score_prompt(prompt, model) → 0–100 score + what's missing",
+        "- enhance_prompt(prompt, model) → AI rewrites it for the model's native language",
+        "- creative_patterns(model) → V3 graph: community A-grade patterns",
+        "- community_benchmark(prompt, model) → V3 graph: how you compare to top scorers",
+        "- my_story() → scoring history, model stats, creative DNA",
+    ]
+    return "\n".join(lines)
+
+
+# ── Entry point ────────────────────────────────────────────────────────────
+
+def main():
+    """
+    Entry point.
+    - DALI_TRANSPORT=http  → streamable-http on PORT (for hosting your own instance)
+    - default             → stdio (local / Claude Code)
+    """
+    transport = os.environ.get("DALI_TRANSPORT", "stdio")
+    if transport == "http":
+        port = int(os.environ.get("PORT", 8080))
+        mcp.run(transport="streamable-http", host="0.0.0.0", port=port)
+    else:
+        mcp.run()
+
+
+if __name__ == "__main__":
+    main()

dali_mcp-0.1.0/docs/best-practices.md

@@ -0,0 +1,234 @@
+# Dali Prompt Best Practices
+
+Model-specific patterns that consistently score A or B. Sourced from Reddit, Discord, and YouTube creators — not official docs.
+
+---
+
+## Video models
+
+### Veo 3.1 (`veo3`)
+
+**The rule: camera move first. Everything else follows.**
+
+```
+Slow dolly push following [subject] as [motion with adverb]. [Location/time].
+[Lighting type] + [lighting quality]. [Style anchor]. No text.
+```
+
+| Do | Don't |
+|----|-------|
+| "Slow dolly push" / "Orbital camera" / "Static wide" | Start with the subject |
+| "gently cascades", "slowly drifts" | Just "cascades", "drifts" |
+| "warm sodium-vapor glow, wet street reflection" | "good lighting" |
+| End with "Cinematic. No text." | Leave mood unanchored |
+| 50–120 words | One-liners |
+
+**Top pattern**: `[Camera]. [Subject + motion adverb]. [Location]. [Lighting: type + quality]. [Style]. No text.`
+
+---
+
+### Seedance 2.0 (`seedance`)
+
+**Write like you're describing a video to a friend. No cinematography jargon.**
+
+```
+A [person description] [does action] in [location]. [Natural light cue].
+Shot in a [UGC/authentic style] way.
+```
+
+| Do | Don't |
+|----|-------|
+| "A woman holds up the product and smiles at camera" | "Cinematic shot of woman" |
+| "Morning light through the window" | "Studio lighting" |
+| "TikTok-style", "authentic feel" | Cinematography terms |
+| 20–60 words | Overly detailed prompts |
+
+**Top pattern**: `[Person + action]. [Natural light]. [Social context: TikTok/Reels/UGC].`
+
+---
+
+### Kling 3 (`kling`)
+
+**Motion amplitude and expression are your biggest levers.**
+
+```
+[Subject with expression]. [Motion with amplitude modifier] in [location].
+[Camera move]. [Mood].
+```
+
+| Do | Don't |
+|----|-------|
+| "subtle sway", "dramatic turn", "gentle tilt" | "moves", "walks" |
+| "warm smile", "focused expression", "wide-eyed" | Skip expression |
+| "Shot 1: ...\nShot 2: ..." for multi-shot | Freeform paragraphs |
+| "Intimate. Warm." as a closing line | No mood anchor |
+
+**Multi-shot pattern**: Label each beat — `Shot 1:`, `Shot 2:`, etc. Kling reads these natively.
+
+---
+
+### Runway Gen 4.5 (`runway`)
+
+**Motion is the noun. Subject is the vehicle.**
+
+```
+[Physical force or motion]. [Subject in that motion]. Camera [tracking move].
+[Location]. [Style].
+```
+
+| Do | Don't |
+|----|-------|
+| Open with a verb or force: "Rain crashes", "Wind tears through", "Light floods" | Static descriptions |
+| Physical forces: gravity, wind, water, pressure | Abstract descriptions of stillness |
+| Camera tracking AFTER establishing motion | Lead with camera |
+| 30–80 words, tight | Long atmospheric paragraphs |
+
+**Rule**: If nothing moves in your prompt, nothing will move in the output.
+
+---
+
+### Wan 2.7 (`wan`)
+
+**Describe the sound. Wan generates native audio from the prompt.**
+
+```
+[Scene description]. [Motion description]. We hear [sound description].
+[Duration pacing]. [Mood].
+```
+
+| Do | Don't |
+|----|-------|
+| "We hear rain pattering softly on leaves" | Skip audio |
+| "Accompanied by a slow jazz piano" | Assume it'll infer audio |
+| "15 seconds", "brief moment", "gradually" | Unspecified duration |
+| Scene → Motion → Sound → Mood order | Random order |
+
+---
+
+### Minimax Hailuo 2 (`minimax`)
+
+**Brackets are required for camera moves. Plain text is ignored.**
+
+```
+[Scene]. [Subject action]. [Emotional beat]. [Pan left] or [Dolly in].
+```
+
+| Do | Don't |
+|----|-------|
+| `[Pan left]`, `[Dolly in]`, `[Close-up]`, `[Tracking shot]` | "pan left", "dolly in" (no brackets) |
+| Multiple brackets in one prompt | One camera move max |
+| Natural language for everything except camera | Use brackets for non-camera text |
+
+**Rule**: `[bracket]` syntax = camera. Natural language = everything else.
+
+---
+
+### Higgsfield (`higgsfield`)
+
+**Name the physics. Describe materials in motion, not motion abstractly.**
+
+```
+[Material in motion: silk draping / hair streaming / water cascading].
+[Character detail]. [Camera movement]. [Style].
+```
+
+| Do | Don't |
+|----|-------|
+| "silk draping", "hair streaming in wind", "water cascading over stone" | "things move naturally" |
+| Name the physics: "cloth simulation", "fluid dynamics", "particle trail" | Vague movement language |
+| Separate character physics from camera movement | Conflate them |
+| Omit "stay still" | "Stay still" creates physics conflicts |
+
+---
+
+## Image models
+
+### Flux Pro (`flux`)
+
+**Write a photography brief, not a description.**
+
+```
+[Camera body], [lens], [aperture]. [Subject front-loaded]. [Lighting: type + quality].
+[Technical style].
+```
+
+| Do | Don't |
+|----|-------|
+| "Sony A7 IV, 85mm f/1.4, f/1.8" | Omit camera specs |
+| "Hasselblad", "Leica M11", "Fujifilm GFX" for premium feel | Generic "professional camera" |
+| 60–150 words with technical specificity | Short vague descriptions |
+| Subject first, then specs, then lighting | Specs first |
+
+---
+
+### Midjourney V8.1 (`midjourney`)
+
+**Keyword list + parameters. Not sentences.**
+
+```
+[subject], [style descriptors], [lighting], [mood], [composition] --ar 16:9 --s 300 --v 8.1 --style raw
+```
+
+| Do | Don't |
+|----|-------|
+| Comma-separated keywords | Full sentences |
+| `--ar 16:9 --s 300 --v 8.1 --style raw` appended | Forget parameters |
+| Style refs: "in the style of [artist]" | Generic "beautiful" |
+| Lighting keywords: "rim lighting", "golden hour", "dramatic side light" | Skip lighting |
+| Remove conversational language | "please make", "I want" |
+
+---
+
+### Ideogram V4 (`ideogram`)
+
+**Text to render MUST be in double quotes inside the prompt.**
+
+```
+[Background/composition context]. [Font style]: "[TEXT IN QUOTES]". [Color palette].
+```
+
+| Do | Don't |
+|----|-------|
+| `"SALE"`, `"Open"`, `"DailyWell"` — exact text in double quotes | "text that says SALE" |
+| Keep each text element under 5 words | Long text strings |
+| Describe font: "bold sans-serif", "neon 3D", "handwritten", "italic serif" | Skip font style |
+| Describe the background + composition | Prompt only the text |
+
+---
+
+### Adobe Firefly 5 (`firefly`)
+
+**Commercial context explicit. No trademarked names.**
+
+```
+[Commercial context: product photography / brand asset]. [Subject precision].
+[Lighting setup]. [Style]. [Color mood].
+```
+
+| Do | Don't |
+|----|-------|
+| "product photography for a health supplement brand" | Vague "photo of a bottle" |
+| "professional", "editorial", "clean studio", "lifestyle" | Generic "high quality" |
+| End with color mood: "Warm amber tones. Premium feel." | Skip mood |
+| Avoid trademarked brands, characters | Name real brands/IPs (defeats IP-safety) |
+
+---
+
+## Universal rules
+
+1. **Score before you generate.** A D-grade prompt wastes the credit. `score_prompt → enhance_prompt → generate`.
+2. **Specificity beats quality adjectives.** "warm sodium-vapor glow" > "beautiful lighting".
+3. **Name the motion, not the result.** "silk draping over her shoulder" > "looks elegant".
+4. **Camera ≠ style.** They're separate signals. Both matter.
+5. **End with an anchor.** A mood/style closing line (even just "Cinematic. No text.") consistently lifts scores.
+
+---
+
+## Contributing
+
+Found a pattern that consistently scores A on a specific model? [Open an issue](https://github.com/Lulu-The-Narwhal/dali-mcp/issues) with:
+- Model name
+- The pattern (ideally as a template)
+- A real prompt + result that scored A
+
+Source it from Reddit, Discord, or YouTube — practitioner evidence beats theory.

dali_mcp-0.1.0/pyproject.toml

@@ -0,0 +1,32 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "dali-mcp"
+version = "0.1.0"
+description = "Dali by Lulu — creative intelligence MCP. Score your prompt before you spend the credit."
+readme = "README.md"
+requires-python = ">=3.10"
+license = { text = "MIT" }
+keywords = ["mcp", "ai", "prompts", "creative", "image", "video", "midjourney", "veo", "higgsfield", "flux", "lulu"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+]
+dependencies = [
+    "fastmcp>=0.1.0",
+    "httpx>=0.27.0",
+]
+
+[project.scripts]
+dali = "dali.server:main"
+
+[project.urls]
+Homepage   = "https://dali.getlulu.dev"
+Repository = "https://github.com/Lulu-The-Narwhal/dali-mcp"
+
+[tool.hatch.build.targets.wheel]
+packages = ["dali"]

dali_mcp-0.1.0/server.json

@@ -0,0 +1,23 @@
+{
+  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
+  "name": "io.github.Lulu-The-Narwhal/dali",
+  "title": "Dali by Lulu",
+  "description": "The prediction MCP that helps you avoid the AI generation tax. Score your prompt 0-100 before generating — so you never waste a credit on a bad prompt again. Supports Veo 3, Seedance, Kling, Runway, Wan, Minimax, Higgsfield, Flux, Midjourney, Ideogram, Firefly.",
+  "version": "0.1.0",
+  "packages": [
+    {
+      "registryType": "pypi",
+      "identifier": "dali-mcp",
+      "version": "0.1.0",
+      "transport": {
+        "type": "stdio"
+      }
+    }
+  ],
+  "remotes": [
+    {
+      "type": "streamable-http",
+      "url": "https://dali.getlulu.dev/mcp"
+    }
+  ]
+}

dali_mcp-0.1.0/skills/dali-creative-flow.md

@@ -0,0 +1,115 @@
+# SKILL: Dali Creative Intelligence — Proactive Prompt Scoring
+
+You have access to the **Dali MCP** (`dali.getlulu.dev`). Use it BEFORE any image or video generation.
+The golden rule: **never spend a generation credit on an unscored prompt.**
+
+## When to activate
+
+Any time the user's intent involves generating, creating, rendering, or making an image or video:
+- Keywords: generate, create, make, draw, render, animate, illustrate, design, produce
+- Media types: image, photo, picture, video, clip, footage, reel, scene, animation
+
+Also activate when a user pastes a raw prompt without asking to score it — score it anyway.
+
+## The Flow
+
+### Step 1: Detect model
+If the user doesn't specify a model, infer from context:
+
+**Video:**
+- Cinematic brand/narrative → `veo3`
+- UGC, TikTok, social → `seedance`
+- Character animation, facial performance → `kling`
+- VFX, motion-first → `runway`
+- Native audio, 4K → `wan`
+- Cinematic storytelling with camera control → `minimax`
+- Physics-driven (cloth, hair, fluid) → `higgsfield`
+
+**Image:**
+- Photorealistic, product photography → `flux`
+- Artistic, editorial, stylized → `midjourney`
+- Typography, logos, text-in-image → `ideogram`
+- IP-safe commercial → `firefly`
+
+If still ambiguous: "Which model are you targeting?" and list: `veo3, seedance, kling, runway, wan, minimax, higgsfield, flux, midjourney, ideogram, firefly`
+
+### Step 2: Score immediately
+Call `score_prompt(prompt=<prompt>, model=<model>)`.
+
+Present concisely:
+- Grade + score: "**Grade C — 58/100**"
+- 2–3 bullet points of what's missing (from `missing` array)
+- One-line verdict
+
+### Step 3: Decision
+
+| Grade | Score | Action |
+|-------|-------|--------|
+| A | 90+ | Confirm and generate |
+| B | 70–89 | Confirm and generate |
+| C | 50–69 | Offer to enhance. If user agrees or doesn't object → enhance |
+| D/F | < 50 | Enhance automatically. Show before/after. Use enhanced prompt. |
+
+### Step 4: Enhance (when triggered)
+Call `enhance_prompt(prompt=<prompt>, model=<model>)`.
+
+The tool returns a **rewrite brief** — not a finished prompt. YOU write the enhanced prompt using it.
+
+The brief contains:
+- `native_language_rules` — how this model thinks
+- `structure_template` — the prompt format to follow
+- `priority_fixes` — what to add, by weight × gap
+- `length_target` — target word count
+- `what_to_add` — specific missing elements
+
+**Your job**: write the enhanced prompt using these rules. Then call `score_prompt` on what you wrote to verify improvement.
+
+Show:
+- Score delta: "8 → 91 (+83 pts)"
+- What changed (key additions from the brief)
+- Enhanced prompt in a code block
+- Score verification result
+
+### Step 5: Generate
+Always use the highest-scoring prompt (your enhanced version, not a raw API field).
+
+## Output format (keep it tight)
+
+```
+Scored: **Grade F — 8/100** for Veo3
+Missing: camera move, motion description, lighting · 8 words only
+Verdict: Generic stock footage guaranteed. Enhancing...
+
+Brief applied:
+  ① camera move first: "Slow orbital push"
+  ② physics: "drop falls", "liquid ripples", "glass refracts"
+  ③ lighting: "warm backlight", "rim-lit edges"
+
+Enhanced prompt:
+```
+"Slow orbital push around a glass serum bottle on white marble. A single amber
+drop falls in extreme slow motion, catching warm backlight. Macro: liquid gold
+ripples outward from impact. Rim-lit edges, soft studio diffusion. Premium,
+clinical. No text."
+```
+
+Re-scored: **Grade A — 91/100** ✓ Safe to generate.
+Ready to generate with this?
+```
+
+## Rules
+
+- Score FIRST, always — before generating anything
+- Don't ask permission to score — just do it
+- Do ask permission to enhance if grade is C (auto-enhance D/F)
+- Never dump raw JSON — translate to human-readable output
+- Lead with grade + missing — not the full dimension breakdown
+- **You write the enhanced prompt from the brief** — don't look for an `enhanced_prompt` field in the response
+- Always re-score your enhanced version to show the delta
+
+## Community tools (bonus)
+
+- Grade C or below → call `community_benchmark(prompt, model)` to show missing A-grade patterns
+- "What makes a good [model] prompt?" → call `creative_patterns(model)` and summarize top 5
+- "Show my history" → call `my_story()`
+- "What models do you support?" → call `list_models()`